You have a bull picked out and a cow in mind. You pull both pedigrees and a name shows up on each side, a few generations back. Same good sire, once through the bull’s line and once through the cow’s. The question at the pen is simple. Is that a problem, and how would you know before you ever turn them out together?
This piece is about answering that question ahead of the mating. A companion guide, reading inbreeding coefficients, explains what an inbreeding coefficient is and what it can and cannot tell you. Read that one first if the term is new. This one is about doing something with it: setting a line you will not cross, and checking a pairing against that line before you breed.
The one number you get to compute yourself
Almost everything on a bull page comes from somewhere else. The American Angus Association runs a breed-wide evaluation every week and hands you the EPDs, the $Values, and the accuracies. No breeder computes those, and neither do we. We read them. That boundary is the whole point of how this operation runs, and it is spelled out in the methodology.
The coefficient of inbreeding, or COI, is the honest exception. It is a pedigree calculation, not a genetic reading, so you can run it yourself from a pedigree with no help from the national evaluation. It uses a method worked out by Sewall Wright long before EPDs existed. It is the one genetic-flavored number a breeder legitimately computes on top of the published Angus data. So it is worth learning to use well.
Linebreeding on purpose, and inbreeding by accident
Inbreeding is not a sin to purge. Mating relatives on purpose, to lock in a good animal’s genes, is called linebreeding, and every settled cow family has some of it. Done with a plan, it is how a herd gets consistent. The danger is not linebreeding. The danger is inbreeding you did not mean to do and did not measure, the kind that creeps in because the same popular AI sire sits behind half the breed and shows up on both sides of a mating without you noticing.
The difference between the two is not the number. It is whether you chose it. A COI does not tell you a mating is good or bad. It tells you how tight you are pulling the line, so a choice you make on purpose stays a choice and does not turn into a surprise.
How much is too much
There is no single number that is the wall for every herd, so this guide will not hand you one. Where the line sits depends on your cattle, your records, and what you are trying to build. Some operations keep new matings under a low single-digit percentage and rarely go past it. Others will accept more inside a cow family they know cold, because they know what is back there and they are concentrating it on purpose.
Two real costs rise as the COI climbs, and both are worth naming. The first is inbreeding depression. As more of the genome doubles up, the traits that quietly erode are the fitness ones: fertility, calf vigor, how long a cow stays in the herd. Growth and carcass numbers often hold up fine. It is the breeding and the staying-alive traits that pay the bill first, and those are the ones you cannot afford to lose. The second cost is the recessive one, and it deserves its own paragraph.
The recessive link
A recessive is a hidden version of a gene. It stays out of sight unless an animal gets two copies of it, one from each parent. A carrier looks completely normal and breeds normal, right up until it meets another carrier of the same thing. That is the mechanism behind the simple-recessive genetic conditions the Association tests for, the ones like AM, NH, CA, and DD.
Here is why the COI matters for this. Inbreeding raises the odds that a calf inherits two copies of the same allele from a shared ancestor. That is true for every gene, including a bad recessive nobody has tested for yet. So a rising COI is a general warning that doubling up is more likely. It is not a substitute for a carrier-by-carrier screen, and a screen is not a substitute for it. The screen catches the named conditions you can test for. The COI flags the general risk across everything, tested or not. You want both. The reference guide covers the carrier math in full in its genetic-conditions section.
Scoring a mating before you make it
Now the useful part. You do not have to breed the cow, wait on a calf, register it, and let it hit the evaluation to find out how inbred it would be. The math that computes a COI for a born animal works just as well on a pairing that does not exist yet. Give it a proposed sire and a proposed dam, and it returns the COI their calf would carry. That is the coancestry of the two parents, and it is exactly the number the future calf would show.
The repository ships this. The inbreeding tool takes a --sire and a --dam and scores the mating ahead of time. Run against the small sample pedigree that ships with it:
inbreeding-coi examples/sample_pedigree.csv --sire C202 --dam C203
Inbreeding coefficient (COI) for C202 x C203 (projected calf): 0.15625
Method: wright_path (Wright's path coefficients)
Pedigree depth used: 4 generations of known ancestry
Contributing paths (common ancestor: sire-chain | dam-chain = contribution):
C201: C202 -> C201 | C203 -> C201 (1+F_A=1.25) = 0.15625
Read what it is telling you. This calf would land at 0.15625, which is 5/32. The whole number comes from one shared ancestor, C201, reached one step up each parent’s line. And C201 is itself inbred, at F=0.25 from a full-sib mating, so it carries extra weight, the 1+F_A=1.25 in the output. Inbreeding stacked on inbreeding compounds. C202 and C203 are just half sibs through C201, and a plain half-sib mating would have guessed a flatter 1/8. The inbred ancestor is what pushes it up to 5/32. None of these are real animals. It is a synthetic pedigree built to show the math working.
Using it as a guardrail
The way to work is to pick your ceiling first, then run candidate matings against it. You are not selecting for a low COI. You do not chase the number toward zero, because that just makes you pass over good matings for no gain. You select on merit, meaning the EPDs and $Values the Association publishes and the parent-average of a proposed pairing, and you use the COI to hold that selection under the line you set. Merit is what you are trying to move. Inbreeding is the line you agree not to cross while moving it.
One caution carries straight over from the other guide. A COI is only as deep as the pedigree behind it. Feed the tool three generations and it reports the inbreeding it can see in three generations, then calls the rest zero. A thin or gapped pedigree makes a mating look safer than it is, which is the wrong direction to be fooled. That is why the tool prints the pedigree depth it used, right there next to the number. A 2% projection on a four-generation pedigree and a 2% on a ten-generation pedigree are not the same claim, and the depth is how you tell them apart.
So the routine is short. Set the ceiling. Before a mating, score the pairing. Read the number next to the depth it came from. Keep it under your line, and let the published Angus numbers decide merit. The COI is the instrument. You are still the one setting the policy.
Related: Reading inbreeding coefficients · Understanding Angus genetic selection · Inbreeding on top of Angus EPDs · Methodology · Back to the reading index
Leave a Reply